The development of dynamic centers of artisanal production in western India during the late colonial period was made possible by complex forms of controlling and utilizing the labour of weaving families. This essay seeks to examine those forms of organization in the handloom industry of the Bombay Presidency between the late nineteenth century and the beginning of World War II. It draws upon existing documentary evidence but also upon numerous interviews with persons connected to the textile industry in the old handloom centers of the region.Footnote 1
Serious discussion of labour in artisanal industry during the nineteenth and twentieth centuries is relatively a recent development. To a great extent, historians working on the colonial period inherited a mindset that simply ignored the subordination of labour to capital in ‘informal’, small-scale, and artisanal enterprise. This mindset had two major sources. First, it stemmed from a Marxist tradition that prompted scholars to look at capitalist relations principally in large-scale enterprises, both in Europe and in South Asia. Second, it had roots in the nationalist, especially the Gandhian, outlook on the South Asian past (a view which itself had strong colonial antecedents). In this perspective, weavers and other artisans were largely actors situated outside capitalism whose livelihoods then became exposed to the disruptive effects of competition with European factories after British rule was established. Until the late 1980s, most scholars who studied handloom weavers during the colonial period kept an almost single-minded focus on the subject of ‘deindustrialization,’ that is, the question of whether or not there was a radical decline in the number of weavers and the volume of cloth produced on handlooms during the nineteenth and early twentieth centuries. While that debate has been an important one, its place of dominance in the literature owes much to the larger symbolic importance of handlooms, as an emblem of the ‘traditional’ and ‘indigenous’ opposed to the forces of colonialism and industrialization.
As a result of such preoccupations, there has been at best a limited recognition of even the existence of labour-capital relations within artisanal production. Little attention has been paid to issues relating to the organization of work, processes of appropriation and exploitation, and workplace conflict, which under other circumstances might be expected to be the central stuff of social history. Ironically, while historians of the eighteenth century and earlier have paid significant attention to relations between cloth producers and merchants,Footnote 2 few scholars have tended to take up similar issues for the colonial period. This shortcoming has had serious implications for our understandings of twentieth century capitalism. Without analyses of labour within artisanal production, there can be only a limited appreciation of the historical processes by which many contemporary forms of industry have taken shape, since much of India's powerloom sector today developed out of the pre-existing handloom sector.Footnote 3
Recent work, particularly the scholarship of Tirthankar Roy, has begun to address this deficiency. Roy devoted a chapter to relations of production in his pioneering study on Indian handlooms and again makes this a theme in his more recent book on ‘traditional’ industry.Footnote 4 In these studies, Roy argues that overall there was a general transition in the organization of labour in India's artisanal industry during the first half of the twentieth century, from the prevalence of “‘independent’ weavers to different kinds and degrees of dependence,”Footnote 5 from ‘custom to contract,’Footnote 6 and from family-based forms of production to the small factory.Footnote 7 In the earlier of these works, Roy argues that weavers increasingly had to submit to some form of ‘tied sale,’ where they were compelled by advance arrangements to provide their product to specific capitalists. ‘Wage contracts,’ in which weavers sold their labour, replaced ‘price contracts,’ in which artisans had been able to negotiate the price of goods sold to merchants. Ultimately, weavers came to form a ‘proletariat,’ which in some cases ‘more and more resembled a class,’ and which acquired a ‘proletarian identity from out-of-caste or familial affiliations.’Footnote 8
Roy thus establishes clearly the importance of labour-capital relations within the handloom industry, in effect demolishing older nationalist understandings. His meticulous attention to the specificity of organizational structures in different places sets a standard for all future investigations of this subject. Yet at the same time, his work would seem to place Indian handlooms too squarely within the familiar master narrative of ‘proletarianization,’ that is, a model of change in which economic actors once located in stable and supportive traditional social matrices become reduced by the all powerful forces of capitalism into (often rootless) individuals who are compelled to exist at the whim of market forces and who ultimately try to cope by developing ties based upon class. Recent research on large-scale industry in South Asia has in effect challenged the value of this teleology by engaging in in-depth analyses of capital-labour relations at the level of the factory.Footnote 9 The fact that the model of proletarianization might seem applicable to small-scale production, I would suggest, may lie in the difficulty of finding source material that would permit similarly close analyses of the internal dynamics of manufacture by the household and workshop. This essay is an effort to examine intensively the organization of the handloom industry in western India, and to reconstruct the processes by which production relations were formed. Labour, I stress here, had important elements of agency in these processes, and capitalist strategies within the workshop were conditioned by a need to cope with and counteract those forms of agency.
The model of proletarianization, I conclude, ultimately proves unsatisfactory in understanding the labour process in handloom production for several reasons. First, by positing the replacement of one kind of ordering principle by a categorically different one (from custom to contract, from family to factory, etc.), it exaggerates the radicalness of change. Many small producers in western India were enmeshed in relations of subordination to, and dependence upon, capitalist actors well before the twentieth century. The major forms of the organization of production that existed in the nineteenth century continued to be found in the twentieth century, though important shifts in the mix of those forms did take place. I concur with Roy that wage employment was growing among weavers and other artisans, though any transition during this period is less dramatic than his account seems to imply. The shape of capital-labour relations among the handlooms remained highly heterogeneous. The importance of mercantile actors from outside weaving communities, for instance, varied from place to place, and also waxed and waned, as businessmen calculated their interests as well as their capacity to impose discipline upon workers in different towns and at different points in time. At any given moment, artisanal production in the Bombay Presidency was characterized by a wide range of structures and practices, from households with one or two looms worked by family labour to larger workshops reliant on hired workers.
Second, the notion of proletarianization suggests a one-sided process in which new forms of organization were simply imposed upon producers. I argue instead that forms of productive relations did change in the twentieth century, but they developed their particular shape out of a complex politics of the workplace and locality,Footnote 10 in which artisan-workers were important participants. Everyday actions by workers, including small acts of resistance at the level of individual workshops, placed limitations upon the kinds of disciplinary orders bosses could establish.Footnote 11 Regional merchants and workshop owners still managed to appropriate most of the profits from cloth sales themselves, reducing most artisans close to the margin of subsistence. But they had to adopt a set of localized mechanisms for overcoming the potential for independence possessed by the weavers, and they often had to accommodate themselves to the cultural practices of their work forces. The role of the family economy, artisanal rhythms of work and leisure, the special forms of artisanal geographic mobility, and the importance of community affiliation all influenced the shape of penetration by merchants and workshop owners in the region. In general the various bosses involved in the small-scale manufacture of cloth established more effective control over the products of handloom weavers than they did over the actual labour processes involved.
Third, the social character of production relations that emerged also fits awkwardly with any notion of proletarianization. Relationships in the twentieth century handloom industry sometimes were partially based upon formal contracts, though I am far from convinced that such contracts became the norm, even in the larger towns of the Bombay Presidency. Non-contractual relationships based upon family, kinship, caste and patronage–social ties that were themselves in the process of construction in the small towns of western India—continued to be central to the maintenance of control and discipline over artisan labour. In some cases, the construction of production ties along these lines was as much a product of the strategies of capitalists seeking to maintain their hold over workers as it was a sign of any kind of stable artisanal tradition. Any movement toward the formation of horizontal ties between workers was complicated by developments that sometimes strengthened the workers' relationship with their bosses and other powerful figures in local society. “Class” proved to be only one of several ordering principles characterizing the politics of workers.
This essay is organized into three parts. In Section I, I examine broad patterns of change and continuity in the organization of production from the mid-nineteenth century up to around 1940. The heart of the analysis lies in Section II, which explores the character of handloom organization and outlines the micropolitics of the workshop. I discuss there three types of productive unit—the “independent” artisan household, the dependent artisan household and the karkhana (larger workshop) – and suggest in a discussion of the latter two types some of the factors that were involved in the formation of production relations. Section III briefly sketches some developments in the collective politics of handloom weavers during the 1920s and 1930s. It focuses on the city of Sholapur, where the growth of Communist trade unions was most pronounced. It demonstrates that even there the politics of weavers and other artisans was rarely driven by a simple contradiction between capital and labour; indeed, the very character of the union movement was conditioned by the ambivalent nature of workers' ties with their employers.
Section I: Change in the Structure of Production
During the nineteenth and early twentieth centuries, handloom production in western India was a changing industry, adapting to the context of international capitalism rather than succumbing to it. In other work I have discussed several important characteristics of these patterns of adaptation:Footnote 12 the specialization of handloom weavers in market niches where mill-made cloth was unable to compete effectively; the specialization of particular locales in particular kinds of products; the increasing reliance of weavers on raw materials, such as cotton yarn and synthetic dyes, that were either imported from Europe or East Asia or that were produced in large factories in India; the concentration of production in small towns, where marketing facilities developed for the disposal of cloth and for the procurement of supplies; and the growing insecurity of weavers' livelihoods as a result of their dependence on markets for cloth that were especially subject to agricultural cycles and on markets for raw materials in which prices vacillated according to international conditions. In the context of the current essay, a particularly important trend in many places was the migration of labour from famine-prone areas of India, where weaving was in marked decline, to the Bombay Presidency. Two main streams of migration developed. The first was the movement of Muslim weavers, known as Momins (or Julahas), from the United Provinces to centers along or near the Agra-Bombay highway, such as Dhulia, Malegaon, Yeola and Bhiwandi. The other was the migration of Hindu Padmasali weavers from the Telangana districts of Hyderabad State to towns in the east of the Bombay Presidency like Sholapur, Ahmednagar and Poona. While both of these sets of migrants appear to have been fairly diverse in background, both in Maharashtra came to forge strong senses of community identity as they settled in specific urban neighborhoods, attempted to gain control over weaving jobs in these localities, and created new social institutions to promote their advancement. In several towns, the Momins and Padmasalis achieved positions of dominance in the weaving profession, replacing artisans from local communities, such as the Koshtis and Maharashtrian Salis.Footnote 13
Many of these developments–-the growing distance of producers from the sources of their raw materials and their markets, their vulnerability to severe annual and cyclical oscillations in demand, the variations in supply and price of raw materials, the widespread displacement of weavers from their homelands and their concentration in larger centers–-bolstered the role of various kinds of small capitalist actors in the handloom industry. Weavers frequently lacked the bargaining power in the yarn or cloth markets necessary to buy or sell on their own in these markets without considerable losses. Often they purchased very small quantities of supplies at a time, sometimes just enough to make one sari. It was impractical for them to obtain yarn directly from Bombay or even from local spinning mills in Sholapur and Hubli. Mastery over the knowledge necessary to obtain good prices for their cloth, moreover, was a skill not easily acquired. Weavers were too far removed from consumers to know personally what the buyers wanted and what they would pay. And because weaving families often needed cash quickly to cover living expenses and yarn purchases, they rarely had the staying power needed to negotiate favourable terms in the market. All of these factors led them to rely heavily on the businessmen who provided yarn and sold cloth, and who in some cases hired them as labourers. The concentration of cloth production in towns put many artisans in close proximity to merchant capitalists, whether dealers in yarn, sellers of cloth, or persons who combined both functions. Or to capture this point even more accurately, populations of weavers and populations of these businessmen grew in conjunction with each other in western Indian towns as the economy of small-scale cloth production expanded.
There was, however, no unilinear path toward some common form of artisan-capitalist relationship. Records indicate at least four general types of relation throughout the period studied here, though lines between these types were blurred. First, there were ‘independent’ weaving households which purchased yarn on their own, and which theoretically were positioned to sell cloth to the highest bidder. Second, there were dependent weaving households, units which relied on merchants for the provision of finance or raw materials, and which had been compelled to enter into arrangements that gave the merchants control over the final product. Third, there were the independent karkhanas, larger workshops containing multiple looms and employing significant amounts of extra-familial labour, where the workshop owners (karkhandars) sold their goods directly in the market. Finally, there were dependent karkhanas, in which small karkhandars did work on contract for other karkhana owners or for cloth merchants. My discussion here will mention all four types, though the second and third will be emphasized.
By the beginning of the twentieth century, purely independent weavers were clearly in the minority. P. N. Mehta, writing around 1909, estimated that fifty per cent of handloom weavers in the Bombay Presidency were persons ‘in debt’, who pledged their looms and their homes to moneylenders or sowkars (shahukars), and who were usually compelled to do the work these bosses demanded (that is, they fell in the second type mentioned above). Another twenty per cent, ‘hopelessly in debt’, worked in the house of the shahukars (that is, they worked in karkhanas, the third type listed above). The twenty five per cent not heavily in debt struggled to maintain their independence (the first type).Footnote 14
It is less clear whether the small number of weavers who could be deemed ‘independent’ in 1909 represented a significant deterioration over the course of the late nineteenth century. Putting out systems and other strong forms of control over labour by merchant-capitalists had often been the norm in regions with substantial commercial manufacture before British rule, especially where the East India Company had come to dominate long-distance trade.Footnote 15 In the Bombay Presidency, there is plenty of evidence from the middle of the nineteenth century of ‘master-weaver’ systems (that is, putting out systems) as well as the existence of small karkhanas with multiple looms worked with wage labour.Footnote 16 In the production of cotton cloth, records indicate many different kinds of actors: independent weavers, master-weavers who ran putting-out systems, and small karkhandars. As early as 1840 one report indicated towns in southern districts of the Bombay Presidency where there were ‘families with some little capital, the owners of several looms, and occasionally carrying on the joint trades of cloth merchant and master weaver, and when this is the case, employed as salesman by their manufacturing neighbors.’Footnote 17 An 1865 account from Belgaum district unambiguously acknowledges the existence of small karkhanas. While most weavers worked on their own account, this report indicated, a few had ‘as many as four or five looms in their houses, worked by hired labour.’Footnote 18 In Ahmednagar around 1880, the District Gazetteer indicated that of the 1200 weaving families, 800 were ‘capitalistic’ and the rest were ‘workmen.’ The capitalists, it was reported, ‘work on their own account and sometimes employed labour.’ The Gazetteer mentions that some of the weavers [seemingly from the capitalist category] were employed by cloth-dealers who supplied them yarn and in turn took ready-made goods from the weavers (i.e., type 2), while others took their cloth ‘every evening to the wholesale traders, and are paid in cash’ (type 1).Footnote 19 Thus, well before 1900, a substantial portion of those who worked as weavers were in some kind of dependent relationship, though these arrangements were quite diverse. Without a baseline clearer than this, it is difficult to tell whether the number of weavers living in dependence had significantly increased by the time of Mehta's study.
The pattern of change in the twentieth century after Mehta's report appears also to be subtle. The evidence seems to suggest less a decrease in the proportion of independent weavers than a partial shift from one kind of tied labour to another, that is, from work in dependent households to work in karkhanas. Roy provides figures from 1940 that 21 per cent of handloom weavers in the Bombay Presidency were ‘independent’, 24 per cent were ‘on contract’ and 54 per cent were ‘factory workers’ (a term indicating employees in karkhanas).Footnote 20 When compared to Mehta's estimates, they would seem to indicate the growing importance of karkhana labour (in types 3 and 4) at the expense of family workers on contract (type 2) but not a significant decline in the total number of independent weavers (type 1).Footnote 21 Since all four forms existed both in 1909 and in 1940, any change over time was a shift in the mix of organizational forms, not the replacement of one type by some new type that had not existed before the late nineteenth century.
For the most part, production structures predominated that allowed appropriation of the weaver's product at very low costs and that simultaneously retained many structural features commonly associated with family-based artisanal production. The ability of capitalists to impose entirely new forms of labour discipline was constrained strongly by two factors. On the one hand, the instability of markets in the kinds of cloth produced by artisans inhibited forms of investment that might have impelled merchants to establish radically transformed divisions of work organization.Footnote 22 On the other, businessmen connected to the handloom industry had in effect to make a series of compromises with the artisans who made up the work force in order to secure their labour. I now turn to a closer examination of particular structures of production, from the artisan household to the karkhana.
Section II: Relations of Production
The Artisan Household
For much of the period studied here, the predominant organizational form of artisan production was the family. In the early twentieth century there were many tens of thousands (perhaps about one hundred thousand) of weaving households in the Presidency. In some respects, these households shared characteristics with what European scholars have termed the ‘proto-industrial’ structure of the eighteenth century in that there was a close association between ‘household production based on the family economy on the one hand and the capitalist organization of trade, putting-out and marketing on the other.’Footnote 23 The focus in theoretical discussions of proto-industrialization on the strategies of families for coping with subsistence clearly is also of some value for understanding the behaviors of weaving households in the Bombay Presidency. Yet the concept of proto-industrialization can not be applied without major modifications. First, most artisan households were not dispersed widely over the countryside but were concentrated in small towns. Second, most tended to specialize in weaving rather than to divide their work time between agriculture and the handloom. Some part-time weavers did exist. In Gujarat, for instance, a number of Dheds (a low-caste category) split their time between weaving and work for wages elsewhere in the countryside.Footnote 24 But the quantities of cloth made by such families appear to have been relatively small (in comparison to those made by weavers with little connection to agriculture),Footnote 25 and most of their production was apparently coarse cloth destined for local circuits of exchange. Third, artisans did not usually participate in the long-distance, international networks that characterized so many producers of eighteenth century Europe or similar figures in nineteenth century Japan. Finally, and most importantly, there is also no reason to believe that artisan households in the Bombay Presidency marked a stage in a necessary evolution towards the factory system, as the term ‘proto-industrial’ implies.Footnote 26 In some centers, no doubt, there was a tendency of the more successful families to develop larger workshops (though rarely modern factories). But in other places, little such tendency existed, and small units managed to maintain or strengthen their role in local cloth production.
In the artisan household, all the work connected to the manufacture of textiles was generally done in and around the home, which functioned as both living quarters and workplace. Weaving families generally owned their own modest houses, their looms and their other tools, though frequently these served as security against debts to moneylenders. Usually the loom itself occupied a central place on the ground floor near the entry way of the home. The space for weaving had to be at least eight feet by sixteen feet so that the warp could be stretched out for several yards behind the loom. Weavers left their position from time to time during the day to release the warp yarn after several yards of cloth had been woven.Footnote 27 In order to maintain proper levels of atmospheric moisture for the yarn, the loom was generally situated in a shallow pit dug into the earthern floor of the home, and the yarn was extended just a few inches above the ground, which was kept wet, sometimes by regular application of fresh cow dung. The weaver generally placed his or her legs into this pit for comfort. When Dr. Hove visited artisanal households in Gujarat during the late eighteenth century, he found the rooms where looms were located shut off from natural light and lit by lamps, which apparently made it easier for the weavers to detect defects in the yarn. More recently, of course, electric lighting has come to be utilized.Footnote 28
The artisan household provided most of the labour it required, and it utilized the efforts of most members of the family. Throughout the region, there appears to have been a strong normative tendency toward the development of a division of labour within the family based upon age and gender. Typically, male adults worked the looms while females, young children and sometimes the elderly helped with preliminary processes: winding yarn onto spindles for the weft, reeling, separating yarn threads in preparation for the warp, sizing, and even dyeing (in cases where dyeing was not performed by separate families).Footnote 29 One observer, highlighting the relations of inequality within the household, pointed out ‘just as the weaver depends upon the sawkar so a class of satellites, often of his own family, depend on the weaver.’ The payments for family labour were usually realized in ‘the form of the price of the finished cloth’ which was paid to the male household head.Footnote 30 Though some of these activities such as warping and sizing came to be carried out by outside specialists in some locales after the 1920s, pirn winding was almost uniformly done by women of the households. In general, young boys became apprentices of their fathers or uncles in their early teenage years before taking up the loom on their own.Footnote 31
Work carried out by women and children was central to the productivity of the family and therefore should hardly be called ‘supplementary’ (a term used in the census and other colonial documents). Warping, for instance, was often done through a labourious process on the streets and alleys in weaving neighbourhoods. In some cases, two poles were fixed on the street and thousands of ends were stretched out between them for every piece of cloth; the yarn was marked every nine yards (the length of a sari).Footnote 32 This activity absorbed a very substantial amount of the family's time. Before 1920 colonial efforts to enhance the productivity of the handlooms by methods that increased the pace of the weaving action were often frustrated by limited changes in the speed of warping; no overall increases in production would take place if more yarn could not be made ready for the loom.Footnote 33 While warping was typically carried out during the day, female family members would do winding late into the night. Women's work appears to have been equally demanding among the producers of silk cloth. There women had to sort, reel and twist the raw silk before the yarn was sent out to specialists for warping; male members of the family did the dyeing and the final processes of weaving.Footnote 34 Women in virtually all artisan households did most of the domestic tasks as well, including cooking, cleaning and caring for children. The household was best able to minimize its costs when unpaid family labour could do all of these jobs and no outside workers had to be hired.
An ideal family size to sustain a weaving household without having to resort to outside labour appears to have been about four to five. Typically, this would mean a male weaver, his wife, one or two children, and perhaps one elderly family member.Footnote 35 When a market existed for its production, such a family could generally maintain a sort of balance between its subsistence needs, its requirements for preliminary labour, and the time available for weaving. Too many non-weaving family members might create consumption pressures that would be hard to meet; too few might mean difficulties in processing sufficient yarn and the necessity of hiring outsiders. Though I am not aware of quantitative evidence that might support this, households may have shaped their demographic behavior–-for instance, by timing births and divisions in joint families–-to strike an equilibrium more effectively.Footnote 36 But no doubt such a balance must have been impossible for many households to maintain due to mortality, variations in fertility patterns, and the maturation of male children. In such instances, the household could make other adjustments. One report suggested that if family size was more than four, ‘it is probable there will be either two weavers or one or more employed in other trades.’ If there was excess non-weaving labor, ‘the women of the family work for any weaver who wants work done and by this distribution of employment the whole community[‘s] inequalities [in terms of needs for weaving labour and needs for preliminary labour] are leveled up.’Footnote 37 Hiring others from outside, working for wages in other artisan households, and even entering the general labour market in the small towns were all strategies that individual families could employ. This obviously meant there could be significant overlap between the weaving household and the workshop based upon wage labour.
Another adaptation families could make was the reallocation of female labour within the household. Interestingly, colonial documents from the Bombay Presidency make little mention of female engagement in weaving itself.Footnote 38 But interviews suggest that women did take up the loom on occasion. Girls grew up in households where exposure to all the processes involved in making cloth was normal, and they often carried their knowledge over to the families into which they married. One can imagine a variety of circumstances in which female skills in weaving might have been utilized: for instance, when a large number of children meant that a family had excess labour for preliminary processes, when illnesses or even death left the family with no available adult male, or when pressures to fill an order created strong incentives for women to weave when looms would otherwise be idle. Though norms that frowned upon weaving by women apparently did exist in some communities, these were hardly so strong as to enforce a completely rigid gender division of labour. In some instances, the skills of women in weaving clearly were critical in the early history of the most entrepreneurial families that expanded their businesses substantially.Footnote 39 Presumably, some degree of flexibility in the gendering of work also could influence male behavior. For instance, family needs might have compelled boys to delay taking up the loom in adolescence or to shift back to yarn processing if the labour available for warping and winding was insufficient. Yet male dominance was generally widespread. In all instances of which I am aware, it was male family heads who received payments from merchants and who thus ultimately controlled family finances.
The need to sell their cloth on a regular basis led weavers into complex relations with mercantile actors. Weavers occasionally appeared in smaller rural bazaars to hawk their wares, but often were unable to compete effectively with larger mercantile actors who carried a richer variety of goods.Footnote 40 Artisan households relied on traders who participated in region-wide commercial networks. Yet these relationships allowed weavers quite differing degrees of control over what they produced. For simplicity sake, we can follow Mehta's categories and classify weaving households into two types—independent and dependent—though we will see that there was no rigid dividing line between the two. After briefly discussing independent households, I will turn to a more detailed treatment of dependent weavers.
“Independent” Households
Most ‘independent’ weavers in the towns of the Marathi-speaking districts (excepting Sholapur) tended to dispose of their textiles through a system known as gujari.Footnote 41 These towns generally had a special cloth market where the shops of adatiyas (brokers) were located. Out-of-town buyers would visit the adatiya quarters whenever they needed supplies. In the late afternoon, usually at a fixed hour, weavers would pass through these markets, carrying saris they had made on their shoulders, on carts, or on bicycles. In some cases they sent their sons with instructions about what prices were acceptable.Footnote 42 Hundreds of weavers would crowd together in the streets, moving from shop to shop as they tried to find buyers. Cloth merchants from out of town would be sitting in the shops of the adatiyas at this time, having communicated their needs to the brokers before the market session began. The brokers, knowing what kinds of qualities and types of saris the different households manufactured, would call weavers whose production matched the buyers' demand into their shops and negotiations would begin. If the weaver was offered a price he could accept, he would sell to the buyer. The adatiya would mark the cloth with his special stamp. The weaver then would take the cloth to another shop of the adatiya, usually located some distance away, where it would be packed and made ready for the buyer when he departed town.
The crucial figure in this marketing process was the adatiya, who was formally an agent of the out-of-town buyer. Adatiyas were key in negotiating the final price obtained by the weavers. When a transaction was made, it was they who paid the weavers, in effect extending credit to the buyers at the same time. The weavers' immediate need for cash meant they were hardly in a position to sell directly to the buyers, who did not expect to repay their debts to adatiyas for up to six months. The adatiya took a commission for arranging the sale and advancing credit.
Such a system obviously placed the weaving family at a disadvantage. It lacked knowledge of what the market outside the town would bear for its cloth, since such knowledge was monopolized by the buyer and the adatiya. In many instances, the weaver came to the market aware that he had to sell his cloth that very day in order to obtain the family's subsistence needs as well as yarn for its next saris. He rarely could wait weeks or even days until prices improved. Morever, he was frequently in competition with other weavers who were also facing the same circumstances. The ‘independent’ weaving family was rarely in a position to secure favourable terms under such conditions.
During the slow seasons of the year, when few out of town buyers were present, the temptation of weavers to sell their goods at reduced rates was especially high. At such times, some of the adatiyas themselves might try to stock cloth, confident they could make larger profits when prices improved. Weavers sometimes asked or begged merchants to take their products when buyers were few. Adatiyas also might themselves become putting-out masters under such conditions, providing yarn to weavers under a contractual arrangement that guaranteed access to the final product. Sometimes the best weavers, whose product was most highly valued, found it easier to make such arrangements than their less fortunate colleagues.Footnote 43
Obviously, then, the independence of the small weaving household was precarious. Such families were in a constant struggle to keep from falling into longterm dependent relations. P.N. Mehta stated of these weavers: ‘their slow and uneconomical system of manufacture, domestic misfortune, bad seasons and famine conditions gradually drag them down from independence to indebtedness and from indebtedness to serfdom under the clutches of sowkars.’Footnote 44 The line between ‘independence’ and ‘dependence’ thus could be thin indeed.
Dependent Households
Dependent households differed little from independent in a number of respects. They too usually controlled their own looms and other tools. For the most part, the artisan family provided most of the labour it required. The division of labour along gender and age lines appears to have been similiar to that in more independent households, and most family members were involved in production. Most importantly, it was the weaving family that determined how work was allocated within the household, not some outside set of actors.
But while such weavers possessed a certain autonomy over labour processes within the home, they lacked much control over the disposal of their goods. With limited resources of their own, they came to rely upon merchant-capitalists who provided them with raw materials or finance, and who expected the final products of the looms in return. During times of low demand, weavers often either borrowed from moneylender-yarn merchants, or they obtained small amounts of work by entering into arrangements that offered the merchants cloth at low prices throughout the year. Through such arrangements, they lost much of their control over their ability to sell their products to the highest bidder.
Dominance over such households was assumed by shahukars (typically translated as moneylender), often traders who had acquired a privileged position in the regional yarn market as a result of their ties to Bombay or up-country spinning mills. In the Deccan and Northern Karnataka, these individuals were usually persons from outside the weaving communities, especially Gujarati and Marwari Banias. Many of them migrated to local towns during the nineteenth or early twentieth centuries to take advantage of the regional trade in handloom goods. With time, they came to exert influence over a number of weaving families by advancing money or yarn to artisans. Evidence from late nineteenth century Guledgud, a major handloom center in Karnataka, suggests that shahukars typically served as master-weavers for 25 to 125 weaving families.Footnote 45
Generally, a kind of putting-out system developed. A weaving household would get an advance of yarn from the cloth dealer-yarn merchant-moneylender. Often this advance would be for a very small amount of raw material, perhaps just enough to make one or two saris. In return, the weaver would agree to provide the merchant with the finished cloth at a given price. The merchant ultimately took upon himself the burden (and profit) of disposing of the final product.
From the perspective of the merchants, there were a number of advantages to these arrangements. The amount of capital a merchant invested was generally rather small, equal only to the cost of the yarn for several saris and any money he had extended to the weaver for performing this task. This mitigated somewhat the risks of participating in the cloth market, or rather, it meant that much of the risk was passed on to the weaving household. If a sudden contraction occurred in the market, the merchant could pull back quickly without losing a large investment in equipment or physical plant. Moreover, he usually negotiated the terms of his agreement with the weaver during the slowest periods of the year, when the artisan family was in its most desperate situation, and when it was ready to engage in extensive self-exploitation to ensure the reproduction of its members and its ability to remain in the weaving business. In short, the system often facilitated a high margin of profit for the merchant during good times, and it limited losses during the bad. This structure thus was one well adapted (from the shahukar's perspective) to the rapidly fluctuating markets in handloom cloth. Merchants were often reluctant to invest in even the simplest form of improved technology that might raise the level of the looms' production but would expose them to increased risks.Footnote 46
But there were also certain drawbacks to this form of organization for the merchants. While able to reduce payments to labour to quite low levels, the shahukars, with little physical presence in the place of manufacture, possessed only a limited ability to exercise any kind of regular supervision over production processes.Footnote 47 The potential was always present for the weaving family to engage in ‘everyday forms of resistance’ that tempered the degree of control the merchant could exert.Footnote 48 Weavers could shortchange the pickcount in the cloth, and then sell any excess raw material or use it to produce cloth on their own account. I interviewed one master-weaver in Surat, who recalled that artisans sometimes used castor oil to increase the weight of the product that they furnished back to him in order to give a false impression of the cloth's quality (that is, of the tightness of the weave), presumably so they could use some of the yarn provided for some other purpose. Alternatively some weavers would cut a yard or two of cloth from the loom and try to sell this on their own. When he paid the weavers for the finished goods, he told me, he had to take into account what he termed weavers' ‘malpractice’. Another merchant similarly mentioned catching weavers who cut yarn from the edge of the cloth, or who increased the weight of the cloth by exposing it to dew or adding water. In such instances, he would normally admonish the weaver, but he had little means of ensuring that these acts did not take place again. ‘This kind of malpractice was going on (all the time),’ he stated, ‘so you had to let it go.’Footnote 49 Such practices were apparently widespread.Footnote 50 The ability of merchants to ensure the quality of their product and the full utilization of all the raw material they provided was thus constrained significantly.Footnote 51
Clearly such behaviors have strong parallels to the European experience before the industrial revolution, where ‘embezzlement’ of raw materials was commonplace and where it appears the practice was widely condoned in the culture of weavers.Footnote 52 In the case of western India, it is impossible to reconstruct the values that sanctioned such acts. To the extent they are mentioned in colonial records, they are critiqued as evidence of the weavers' lack of commitment to capitalist discipline, a portrayal no doubt far removed from the perspective of the actual artisans. The general triumph of the businessman's ethos in contemporary weaving locales means such attitudes are unlikely to surface in contemporary interviews among artisans themselves. In one case, I talked with a weaver who expressed general dissatisfaction with his relationship to his employer fifty years earlier, but who gave few particulars of the actual sources of tension. A local man present told me later that the weaver's reluctance to mention specifics probably stemmed from the fact that his employer had caught him cheating. Not surprisingly, merchants were far more open in discussing the practices of those who worked for them than were employees in acknowledging engagement in such behaviors.
To a great extent, weavers also preserved control over the rhythms of their day, and generally refused to submit to attempts to regulate their work and leisure hours. As one observer remarked of Indian artisans, ‘[the weaver] is a fairly hard-working individual, . . if it were not so he would be starving, but he is accustomed to work at his own time and in his own home, and the regular hours obtaining in a factory are distasteful to him.’Footnote 53 In the Madras Presidency, attempts by the Industries Department to set up government-sponsored handloom factories on a more ‘rational’ basis generally failed because ‘caste weavers’ would not accept a work day governed by the time clock. There, officials complained, the artisans resisted any effort to impose ‘discipline.’Footnote 54 No such attempts seem to have been made by the Bombay government, perhaps because the futility of such an effort was readily apparent. Writing in the 1930s, N.M. Joshi found both independent and dependent weavers in the Bombay Deccan working on loose timetables of their own making that sometimes stretched from early morning to very late at night but included breaks for smoking and visits to the market.Footnote 55 It was, after all, mainly ‘the flavour of independence and ease’ associated with the weaving household that led artisans to persist in a trade that could barely sustain familial subsistence.Footnote 56 Merchants were sometimes unable to coerce greater production during busy times, when the opportunities for profit were at their highest.
This partial autonomy of artisan families could frustrate the objectives of the shahukars in other ways. Weavers sometimes would simply break the exclusive relationships they had with their shahukars by selling to some other merchant. P.N. Mehta suggested that such efforts were commonplace: ‘. . . In good times when the demand for manufactured articles is brisk . . . [they] are often tempted to get a loan from some other person and thus try to deal with more than one Sowkar on the sly.’Footnote 57 A merchant in Surat reported to me that weavers sometimes cut from the loom cloth they were manufacturing for him and began doing work for some other merchant during busy times of the year, when other dealers might offer higher prices for commissioned cloth. He felt compelled to visit weavers' homes regularly to make sure his work was being done on their looms. Such practices may have become especially extensive during times of crisis. During the famine of 1896–7, for instance, weavers in Guledgud, worried they were not going to receive future advances from local Marwari shahukars, simply began taking cloth made from yarn provided by the merchants directly to the local market.Footnote 58 The fact that markets existed in raw materials pilfered by weavers and in cloth made from such materials is an indication that the merchants themselves hardly constituted an effective common front against such practices.
Finally, weavers could simply decide to leave town when the pressures upon them became too great.Footnote 59 A weaver in Sholapur pointed out that artisans who were heavily in debt to a sowcar, and who could not find an alternative patron to assume the debt, in some cases would just flee back to their places of origin in rural Andhra with their dismantled looms. A relative of his did just that on one occasion. Perhaps more than any other occupational category, the weaving population was highly mobile, and emigration was a regular resort.Footnote 60 Preventing workers from leaving or working for other shahukars when demand was highest was a significant concern of many master-weavers.
The continuing importance of community, too, placed constraints on the ability to control artisans. In most weaving centers, the profession was monopolized by members of one or two castes. These collectivities sometimes made strenuous efforts to prevent the skills of weaving from being disseminated more widely. L. V. Tikekar, who set up a weaving factory in Sholapur at the beginning of the twentieth century, reported in testimony before the Bombay Provincial Banking Enquiry in 1929 that his initial efforts to start a handloom enterprise were completely undermined by his inability to recruit artisans from local weaving castes, who felt that their skill would be stolen. Eventually he was compelled to recruit unskilled labourers, whom it took up to two years to teach the kinds of weaving required in his workshop.Footnote 61 Merchants unwilling to train their own work forces had no other option than to rely on labour from the already existing weaving communities, and there was little possibility of developing a ‘free’ labour market unencumbered by caste affiliations. It was reported in 1882 that the Momins of Bhiwandi had such ‘a strong class feeling,’ that they joined together ‘so staunchly in thwarting efforts of court officers that defaulting debtors are seldom caught.’Footnote 62 Most of their debts were probably to Marwari shahukars in their locality.
Pilfering yarn, selling cloth on the side, failing to respond when merchants sought to speed up production, absconding with advances and the like were generally small-scale political strategems designed to give the weaving household some extra breathing space in its efforts to secure its subsistence, leisure and freedom. While these acts may have been sanctioned by larger community norms, they did not assume a collective character. I am not aware of a single incident of joint action by household producers against cloth dealers in western India during the colonial period. ‘Everyday resistance’ was generally an tactic of individual families utilized against individual merchants. This is perhaps not surprising in an environment where production was dominated by a large number of small bosses exercising a loose control over thousands of households dispersed over the landscape of western India. Collective protest seems to have been reserved for persons deemed to be oppressing the community as a whole, such as grain dealers or agents of the state.
But the potential for small acts of resistance was substantial enough that merchants often had to devise their own methods to prevent these behaviors from making serious inroads into their profits. Most importantly they fostered relations of indebtedness and dependence by providing loans on top of the advances of yarn or cash they made available for production itself. Unlike the cooperative societies sponsored by the colonial regime after 1905, shahukars were willing to provide loans for ‘non-productive’ purposes, including wedding and festival expenses of the artisans. In some cases, they charged very little interest on such loans; in other cases they levied very substantial rates, as much as 24% or higher. But more important than the earnings that such lending provided was the improved potential for control over the products of the looms that it made possible. Through debt, the shahukars cultivated relationships of personal obligation that artisan-clients often found difficult to break. When artisans threatened to strike out on their own, for instance, by becoming members of cooperative societies, the moneylender-weavers would press for repayment of debts or claim forfeiture of the weavers' homes. Unless the artisan had somehow been able to accumulate some savings he was unable to break away. Once a debt was contracted, concluded one source, ‘the weaver is likely to remain for ever in a state of want and dependence.’Footnote 63
Manipulation of the methods of repayment of advances could also serve as a means of securing control over the weaver's product. One especially detailed example of such methods comes in 1929 from Gokak, a small weaving town in Belgaum district. There 250 of the 300 weaving families in the town were under the control of the local shahukars, who provided them yarn purchased from the one Marwari yarnseller in the locality; the remainder were independent artisans who relied on their own capital or on loans from a fledgling cooperative society. While advancing yarn, the shahukars, with the weavers' knowledge, regularly debited the account for two or three times the value of the yarn actually provided. When the weavers returned the cloth, they would be credited with an amount similarly exaggerated, thus erasing the yarn debt on paper. The purpose of this accounting procedure appears to have been to compel the weavers to deliver the cloth to the dealer who had furnished them the yarn in the first place. Otherwise they risked being held responsible for the total amount falsely debited. Local shahukars showed little reluctance to take to court those weavers who proved refractory in meeting their financial obligations. In Gokak, too, weavers also borrowed heavily from the shahukars, mortgaging their houses to them, becoming permanently indebted in some cases.Footnote 64 Apparently, local weavers felt they had no option but to accept these arrangements if they hoped to get work.
Such systems of payments, I would argue, can best be understood as forms of response to the weavers' attempts (or anticipated attempts) to elude contractual and oral agreements, that is, as methods of counteracting the weavers' resistances. No doubt because of actual past experiences, the merchants feared that artisan households would either fail to return finished cloth promptly or that they might use yarn advanced to them in order to become sellers able to bargain for better prices in the market. During the twentieth century, the possibility that local cooperative societies might attract these weavers also became a concern in some places. By creating obligations greater than those required for everyday production needs, the merchants ensured their command over the products of the looms in a context where they were unable to dictate the forms of economic activity within the artisan family.
In short, merchants lacked the ability to exert regular supervision and discipline over weaving households, and had to accept a certain degree of independence on the part of even the most dependent of weavers. They instead tried to develop forms of control that would not involve direct intervention in the production process itself. Often these methods involved establishing vertical ties with their workers. These ties sometimes came to constitute the weavers' most important social relationships, overriding horizontal bonds that might have united them with others positioned similarly in the economic hierarchy.
The Karkhanas
Surprisingly, many of these same general features of artisanal practice were reproduced in the larger workshops (the karkhanas, types 3 and 4 discussed in Section I). At first glance, the karkhanas seem to represent purely a system of factory organization based on wage employment, places where supervision and discipline would have logically been much easier to accomplish. Most karkhanas were small units containing four to ten looms in which the loom owner employed much of the labour he needed from outside his family. In some cases, a few big karkhanas with fifty or more looms emerged. In the 1920s, the Irabatti family owned three or four hundred looms, though it is unlikely these were all located in the same building.Footnote 65
We have already seen that some workshops hiring labour existed in western India during the middle of the nineteenth century, though it is difficult to tell how common these were. Karkhanas clearly had become a signficant organizational form in most of the weaving towns by the century's end. R.G. Kakade suggests that the first karkhana in Sholapur, with about ten looms, may have been started by a Padmasali weaver around 1880.Footnote 66 A survey of handloom weaving establishments in the town of Guledgud during the famine of 1896–7 reveals that more than half the artisan firms surveyed contained at least four looms, with some even possessing as many as ten to fifteen. While some of these units may have been staffed by family members, wage labour was no doubt employed in others.Footnote 67 The size of one weaving firm was given as twenty-nine people, suggesting either that labourers were counted as family members or that large extended families functioned as a work unit.Footnote 68
Still, larger workshops employing workers on a wage basis undoubtedly became more prevalent during the first few decades of the twentieth century. The expansion of the karkhanas was particularly associated with the increases in the migration of skilled artisans discussed earlier in this essay. Thousands of Padmasali and Momin weavers from Telengana and the eastern United Provinces were abandoning household employment in their home regions for the Bombay Presidency. Often they found employment with weaver-capitalists who had successfully saved small amounts of money and who had purchased extra looms. By the 1930s, more weavers of Sholapur worked in karkhanas than in household firms, and there were a number of firms with more than forty looms.Footnote 69 There is less evidence of truly big firms in Momin-dominated weaving places like Bhiwandi, Malegaon and Dhule but there, too, the majority of weavers were recent migrants who found employment in workplaces with multiple looms.Footnote 70 The association between migration and the karkhana system alerts us to the limitations of seeing the development of larger workshops simply as a product of initiatives taken by bosses to gain more effective control over workers.Footnote 71 Rather, we must also recognize the processes of dislocation in labour catchment regions that released numbers of skilled but loomless weavers to the small towns of western India.Footnote 72 Without the availability of this pool of weavers, workshop owners might have had to devise very different kinds of arrangements for mobilizing labour. The character of productive organizations, in other words, was as much shaped by the needs of workers as it was by the small capitalists.
The vast majority of entrepreneurs who founded karkhanas were themselves drawn from weaving communities. Most had once been weavers and still retained a high degree of weaving skill. Some of them successfully moved into the manufacture of higher quality saris that offered prospects for better profits, then slowly built up the scale of their operations so they could satisfy the growing demand for their product. A number actually became cloth dealers in their own right by opening shops in the weaving towns. In a few cases, they competed effectively in the cloth business with Marwari and Gujarati Banias, who previously had dominated the textile trade. In the smaller karkhanas, owners, their wives and their children continued to do weaving and other tasks alongside hired labour. But if a firm became large enough, members of the weaving family gradually withdrew from the actual weaving process, and began to engage full-time in supervisory and marketing responsibilities. In a few cases, a brother might even take up government employment or other salaried work. Wives of karkhandars too sometimes assumed a supervisory role over women who did the work of winding.
Karkhandars were not always independent businessmen ‘free’ to sell their own product. Many workshop owners who hired labour were themselves bound by dependent relations with particular cloth merchants. In Sholapur, a large number of karkhandars with multiple looms were known as asamis, or out-workers. Under contract or bound by advances of yarn, these asamis were obligated to provide all the cloth they produced to specific clothdealers, most of whom were karkhandars of Padmasali background. These larger karkhandars drew upon the labour of asami workshops to manufacture saris, particularly during brisk periods when their own establishments could not keep up with demand. Some asami firms were simply weaving households using family labour. But in a sample collected by Kakade in the 1930s, half of the owners of “factories” (workshops with ten or more employees) were in some kind of dependent position and there was at least one owner with twenty six looms in this situation.Footnote 73 Kakade found several gradations of relationships between “independent” and dependent karkhandars; clearly many weavers moved between these two statuses over time.Footnote 74 A system similar to this existed in Ahmednagar.Footnote 75 In general asamis were expected to supply all of their cloth to their karkhandars. But karkhandars did not impose control over the internal workings of the asami firms. ‘The only freedom the out-workers enjoy,’ stated Joshi, ‘is the absence of any time restrictions on their working day and the avoidance of factory conditions of work.’Footnote 76 In short, the relationship between asamis and larger karkhandars was quite similar to that between dependent artisan households and shahukars described in the previous section.
Associated with the increased size of handloom workshops in some cases were changes in the division of labour. In Malegaon, Sholapur, and Sangamner, for instance, there was a tendency by the 1920s to contract out warping work to firms specializing in this task, some of which had introduced new labour-saving techniques that had freed owners from the need to hire female labour for these purposes.Footnote 77 Apparently the savings of these methods were extensive. But at the same time, some weavers in the same centers began to house dyeing processes within their workshops, eliminating their reliance on dyeing specialists and cutting their costs.Footnote 78 Organizational changes appear to have been less profound in some other weaving towns. In all cases, some of the preliminary tasks done in the artisan home continued to be performed in the larger workshops. Pirn-winding, for instance, seems to have been universally carried out by women within the karkhanas.
In essence, this form of organization allowed the karkhandars to avoid some of the risks of participating in an unstable commercial environment that might have existed if they had built more complex forms of industrial establishment. When demand was low, the workshops could simply reduce the number of looms doing weaving (or cut back on providing out-work to asamis), thus releasing some of their workers to unemployment. When demand picked up again, they might rehire labor to meet the new need. Some forms of weaving establishment appear to have been impermanent entities, which were in effect dismantled and then reassembled as market conditions dictated. Joshi found in his study that the karkhanas in Sangamner and Ahmednagar owned by Padmasalis had become entirely seasonal by the early 1920s. Many were open only between December and April, when demand for cloth was at its highest. During this period, they hired Padmasali workers, perhaps weavers who tried to eke out subsistence through household production at other times of the year. These seasonal karkhanas were able to outcompete ‘the permanent establishments generally owned by Deccani weavers, Salis, Koshtis, Khatris, etc.’ because their costs were lower.Footnote 79
Though wage employment undeniably existed in the karkhanas, artisanal practices continued to permeate their social organization. For instance, the artisan family often retained much of its integrity as a productive unit as well as a unit of consumption. In many cases, the karkhandar would hire the family as a whole, agreeing to employ women members to perform preliminary processes. The women thus continued to carry on with a customary sphere of economic activity within artisan households. A sample of 551 female workers in the handloom industry of Sholapur conducted by Kakade found 203 belonged to the karkhandars' families and 348 belonged to the wage-workers' families; there seems to have been no employment of women whatsoever without some kin connection to males in the establishment.Footnote 80 In some instances, a sort of collective family wage was paid for the combined work of different family members, presumably to the male head of the family.Footnote 81 In the 1930s a number of women in Sholapur, perhaps displaced by new processes of warping, began to take up wrapping bidis (cigarettes), but they generally performed this job as outworkers, and they often did this task alongside their husbands, children, and other family members in the karkhanas.Footnote 82 A few women may have become weavers, but the profession remained dominated by males; there thus was much continuity in the gender division of labour. Children also took work in the karkhanas, though not always in places where their parents were employed. A number of boys moved from workshop to workshop as they attempted to gain the skills of weaving. These earned only meager wages as assistants, usually after a period of six months to a year in which they received only room and board.Footnote 83
Despite the larger scale of operation, labour within these karkhanas retained some of the same kinds of independence that it possessed in the artisan households. This manifested itself primarily in three forms. First, the weavers carried into the karkhanas artisanal approaches toward time, work and leisure, in effect resisting any attempt to be regulated tightly by the time clock and the calendar. Observers of the handloom industry in Sholapur reported that it was common for weavers to leave the karkhana during the day to smoke chilim or chew tobacco, then to work late into the night. Workers generally stopped early on market days, and often took time off whenever some important occasion required.Footnote 84 During periods of the year when demand was low, such an approach to time may not have concerned workshop owners, who paid by the piece rather than by the hour. But it could cause disgruntlement among karkhandars who had committed themselves to provide large amounts of cloth during busy seasons. The Sholapur Handloom Enquiry of 1948 reported employer complaints that ‘even when the raw materials are made available, the weavers do not work regularly and the hours of effective work do not on an average exceed five or six a day.’ The Enquiry concluded that these claims were exaggerated, but went on to state ‘we cannot deny the fact with all the atmosphere of freedom and leisure the labourer would always find it difficult to put in sustained work for a regular number of hours. We can not expect regular work when the place of work is also the place of residence.’Footnote 85 Obviously the language of the report reflected the perspective of employers, who had an interest in establishing tighter forms of control and discipline over their work forces and who thus were reluctant to acknowledge the ways weavers spread their work over more loosely structured days, weeks and years. Older employers I interviewed in several different towns mentioned that they had to accept such practices or they would risk losing their workers. These interviews also suggest that the practices continued even in the early years after powerlooms were introduced in some workshops.
Second, labourers remained a highly mobile set of people who retained strong ties with their home districts. Employees in the workshops of Sholapur, for instance, traveled back to their native places in Andhra Pradesh up to three times a year. In some cases, these trips allowed the workers to cope with slow periods of karkhana business. Costs of living in rural areas were lower than in the weaving towns, and out-of-work weavers could hope to find temporary employment in the countryside to supplement their incomes.Footnote 86 But in other instances, social purposes were paramount. Weavers visited their villages to tend to property there, to attend weddings, to visit relatives and small children, and to pay homage to their village deities. They sometimes could leave when karkhandars in Sholapur needed their labour most.Footnote 87 From the perspective of employers, they often were, to use Arjan de Haan's term, rather ‘unsettled settlers.’Footnote 88
Even where labourers were long-term residents of a town without rural connections, however, they possessed potential mobility in another sense. That is, they could seek to transfer to other owners who promised higher wages or more steady employment. Interviews conducted in several towns with both karkhandars and labourers suggest that a significant segment of the weaving population stayed with any single employer only a year or two before moving on. One owner in Surat referred to workers of a slightly later time as being like ‘pigeons,’ always flying from one employer to another. In every town, elderly workers expressed that they had been aware of wages paid in other karkhanas, and that they had been ready to leave (or had threatened to leave) when some other local establishment offered higher pay. Again, such behaviors may not have proved much of a concern during slow times of the year, but they could trouble employers seeking to meet heavy upsurges of demand during peak business periods. Keeping these forms of mobility in check, expressed as the value of promoting ‘the workers' happiness,’ was critical to owners.
Finally, weavers exerted a great deal of leverage in the recruitment process. Perhaps because karkhandars could not offer regular or remunerative employment throughout the year, owners were unable to entice significant numbers of persons from outside weaving communities to enter the industry and thereby enlarge the pool of available labour. During busy times, karkhandars competed with each other to attract weavers to their workshops. Workshop owners from Surat reported that they sometimes tried to woo skilled weavers away from their rivals over drinks in the neighbourhood toddy shop. But most commonly, they relied on their own connections or those of their existing workers to find labour. The majority of the weavers I interviewed stated that they had obtained their positions through a relative or friend already employed in the same workshop. One worker mentioned having been delegated by his boss to return to his home village to bring back a couple of weavers.Footnote 89 Karkhandars generally hired from among the caste fellows of their workers (who in turn were from their own castes).Footnote 90 Such practices certainly had some advantages to the weaver capitalists. According to one report, they rendered the possibility of loss by the absconding of ‘fradulent weavers’ far less likely. ‘The society of caste and trade honor’, this source claimed, ‘would suffice to make the risk negligible.’Footnote 91 But they also made bosses especially reliant on the social networks of their employees.
Karkhandars thus were not able to create out of thin air precisely the kind of work force that they might have wished for. Because the potential independence of their employees was so great, they often attempted to develop mechanisms to tie labour to their weaving establishment, measures that were similar in some respects to the means merchant-capitalists developed for creating clients among the weaving households. These measures may not have reflected a calculated strategy consciously conceived by each employer individually. Rather, it is more likely that they were developed as the result of the experiences of hundreds of owners seeking to overcome workers' actions that had the potential to hurt their businesses, and with time they became engrained in the culture of the workplace. By the early twentieth century, they had become part of a set of mutual expectations on the part of employer and employee alike. When employers sought to structure relations outside of these expectations, they could find themselves faced with considerable resistance.
The most important of these mechanisms in many localities was a system of advances. Karkhana owners in Sholapur offered virtually all of their labourers a significant, interest free advance at the moment employment began, sometimes insisting that a bond of service be signed at the same time. Without such advances, the owners simply could not hope to get access to the skilled labour they needed, as L. V. Tikekar discovered at the very beginning of the twentieth century. Recalling this situation later, he commented: ‘Their [the workers’] way of doing things was quite a different thing from others. If I wanted to engage a weaver . . . . in order to simply engage his services, . . . I had to pay the wages before he did a certain amount of work. This was the kind of thing then.Footnote 92 The practice of advances persisted into the post-independence period. Karkhandars also made loans on the occasion of marriages within the labourers' families and at the time of religious festivals. Finally they sometimes provided advances for ordinary expenditures, such as the purchase of food and clothing. In Sholapur, these loans may have sometimes been paid in the money of Hyderabad State, where most of the Padmasali weavers originated, rather than in the currency of British India, where the karkhanas themselves were located.Footnote 93 Clearly the workers expected to remit the money back to their home villages or to finance further family members to move to weaving centers in the Bombay Presidency. In some cases, a non-weaving Padmasali told me, remittances were used to pay off debts to village moneylenders.Footnote 94
The advance system was not universal for workers in some other centers, but employers everywhere needed to be willing to offer their most skilled and trusted weavers such loans. Karkhandars in Surat, for instance, indicated that they gave advances to those weavers whom they knew best (and whom, no doubt, they wanted to keep the most). The possible consequence of failing to do so is illustrated by one weaver who told me that he once left an employer after many years largely because he was refused a loan. Joshi found that in Ahmednagar and Sangamner advances ranging from Rs. 20 to Rs. 100 were often given to temporary workers, who were thereby ‘maintained as an emergency labour force by an agreement between the weavers and the masters according to which they were given preference by the karkhandars whenever extra labour was required.’Footnote 95
Advances served to solidify the relationship between weaver-capitalist and labourer. In effect the karkhandar fashioned himself not just as an employer but as a patron of his workforce, an individual whose claims over labour were based upon social as well as contractual obligations. The karkhandars took regular installments out of the sums they paid to their workers, sometimes from twenty to twenty-five per cent of total wages, though these amounts were often negotiated at each pay period and could be reduced significantly. Advances of course came with a condition: ‘that the debtor weavers should not give up the employment in their karkhanas unless and until the debts due by them were completely paid off.’Footnote 96
Few labourers managed to free themselves entirely from their obligations to their employers. In some cases, relationships of debt clientage were passed down to the next generation. Indebted weavers wishing to leave a karkhana usually had to persuade their new employer to pay off their advance. In such cases, they generally transferred their loans and their allegiances to the new master-weaver.Footnote 97 No doubt, the need to make such arrangements constrained the impulse to leave on short notice during peak business periods, and it allowed karkhandars to keep payments low with a much reduced fear of losing workers.Footnote 98 In sum, debt appears to have been almost as important a mechanism of control of the karkhandar over the worker in the workshop as it was of the shahukar over the small artisan household, despite the formal existence of wage employment in the former.
The provision of housing to labourers' families could provide another important means of exerting leverage over the workforce. In Sholapur, weavers and their children lived in the karkhana itself, eating and sleeping besides the looms that supplied their sustenance. Few could afford to purchase their own housing, a necessary step if one were to strike out completely on one's own. From the vantage point of the employers, this arrangement may have allowed closer supervision of work. Weavers who traveled to their home districts were generally expected to leave their families behind in the karkhana. In effect these family members became ‘hostages’ (the term is Joshi's) whose presence prevented the weaver from absconding without repaying his advances. In short, the development of wage labour in the karkhanas did not always bring about the delinking of work space and living space that one associates with the transition from household to factory production. From the perspective of the workers, the provision of housing in the karkhana, too, became an expectation of employment in Sholapur.Footnote 99
There were other means of ensuring observance of wage arrangements, whether based upon contract or not. In Sholapur during the 1930s, employers provided workers who made at least four saris a week a significant amount of tobacco for smoking during non-working hours. This provided a subtle pressure on labourers to maintain a certain level of productivity during busy seasons when the employers' ability to compel regular work hours was otherwise limited. During slack times of the year, when production remained below four saris, there was no need to pay the allowance.Footnote 100 But weavers could potentially refuse to work for those who did not furnish such an allowance when the demands for their labour was greatest.
Different forms of social patronage were prevalent in different localities. Sponsorship of communal feasting and drinking appears to have been common in many places. In Surat, karkhandars commonly held feasts at the temple of the Khatri goddess or kuldevi during the monsoon, when there was little weaving work available. One ex-owner told me that his firm sponsored four out of the five days of feasting on such occasions; the workers collectively were responsible for funding the remaining day. ‘If a sheth wanted workers to work,’ he stated, ‘it was important to observe this.’ A Bania capitalist whose family employed Khatri weavers acknowledged that his firm too supported such occasions. Asked why his father would sponsor such feasting though not a worshipper of the kuldevi, he commented that ‘we learned from the Khatris [i.e., Khatri karkhandars]’, clearly admitting a strategic element in this behaviour.
All these mechanisms allowed the merchants to compensate to a great extent for their inability to coerce workers into submitting to what one might call fully capitalist forms of work discipline. As with the weaving households examined in the previous section, what was accomplished was less the ability to dictate the shape of the labour process than it was to secure final control over the cloth produced. In Marxian terminology, this form of organization reflected the ‘formal’ subsumption of labour to capital, in which much of the handicraft character of production is maintained, rather than ‘real’ subordination of labour under capital, in which capital exerts full control over the manufacturing process.Footnote 101
Thus a multiplicity of structures–independent weaving families, weavers working for merchants on a contract basis, labourers employed in small handloom factories and even (by the 1920s) a few powerloom factories–managed to coexist in the Bombay Presidency. Changes in the larger economy had influenced all handloom weavers, but these had not determined any particular form of production relations. The structure of manufacture in the handloom industry was instead shaped by the micropolitics of the artisan household and workshop. Dependent relations of various sorts had become predominant. But even in establishments based primarily upon wage employment, these relations always coexisted with forms of control that reflected an accommodation to the culture and social organization of an artisanal labour force.
Section III: Workers' Identity and Consciousness
The structuring of the workplace undoubtedly affected the kinds of consciousness and identity that labourers in the karkhanas were developing in the early twentieth century. These kinds of self-conception, however, were not evolving in some straightforward manner toward class consciousness and a proletarian identity. Rather they appear to have been a product of the multiple lines of affiliation that had taken shape in association with the complex organizational forms of the handloom industry. The spread of wage labour certainly promoted new kinds of solidarity between those in similar positions. But these horizontal allegiances criss-crossed with vertical ties between workers and their bosses, ties based upon the forms of social patronage and economic dependence I have discussed above. In some instances, kin, neighbourhood and caste coincided with, and reinforced, horizontal linkages. But these same social institutions could also bolster vertical ties with bosses, who could be social leaders in weaver localities, and, in the case of karkhanas, were more often than not caste fellows, even kinsfolk, of their employees.
Evidence concerning weavers' consciousness is scanty and problematic. There are two main kinds of sources for reconstructing this consciousness. First, there are the recollections of old weavers of their relations with their bosses, which can be recovered in oral interviews. These recollections do have shortcomings as evidence. There are now relatively few weavers remaining who were adults before 1940, and recollections about later periods (the 40s and 1950s), when handlooms were abandoned in favour of powerlooms in many of the weaving centers, are inevitably sharper than those for earlier periods. In addition, current memories are strongly coloured by more recent experiences, such as the intense class conflicts of the 1950s and 1960s. Some workers tend to regard earlier times with an uncritical nostalgia.
Workers, however, provide no single version of the past.Footnote 102 Some concur with owners' recollections that strong bonds existed between employer and employee. Production relations in those days, I often heard from workers and karkhandars alike, ‘were just like familial relations.’ Metaphors of father-son or even brotherly ties were sometimes evoked. Some recalled the many patronage services provided by bosses. But in other cases, labourers stressed their conflicts with owners, and remembered incidents of unfair treatment, physical beatings and sexual harassment. Some recollections suggested weaknesses in the solidarity of workers at the time. One jari weaver, clearly remembering with pain some personal mistreatment, indicated that he had to suffer by himself. If there had been a union at the time, he lamented, ‘I would have had to join.’ A few individuals stressed the divided loyalties of workers in their firms, mentioning that the strong ties of a few employees to owners inhibited collective activity by the workers in their unit as a whole. And criticisms of bosses commonly included the charge that they did not listen fairly to complaints against other workers.Footnote 103
The second kind of source on consciousness concerns collective action, particularly that associated with trade unions. Here the information is sporadic. In most weaving towns, there appears to have been very little strike or trade union activity until the late 1940s when political parties began to compete very aggressively for workers' votes.Footnote 104 In Bhiwandi, there is mention of a strike for an increase in wages by two hundred weavers in the handloom workshop of Abdul Samad Sheth in the early 1930s, though reports of this occasion suggest it was settled with no repercussions in terms of extended trade union organization.Footnote 105 Though Samad Sheth's workshop became the focus of considerable strike activity after independence, there appears to have been little trade union activity there or elsewhere in Bhiwandi through the remainder of the 1930s.
One exception to this general pattern was Sholapur. There Communist unions formed in the 1930s, and retained a presence throughout the colonial period. To a great extent, this development was an outgrowth of collective activity in the Sholapur mills, where Communists had developed a strong following among Padmasali weavers.Footnote 106 The foothold gained by the communists among the Padmasali mill workers appears to have been essential to the quick spread of unions in the handloom sector as well as the bidi industry (where women of the community provided much of the labour). Educated Padmasali youth such as V.R. Madur, inspired by the local leaders of the party, formed a handloom weavers' trade union, the Lal Bavta Hatmag Kamdar Union, to fight for labourers' interests. It rapidly gained membership among local weavers.Footnote 107 In 1936 the Congress Party also started a union for handloom weavers that made efforts to settle disputes, though this appears to have been far less strong. In the late depression context of severe wage cuts and employer attempts to withdraw traditional practices such as the tobacco allowance, the Communists fought a series of strikes in different individual workshops. In 1940 there was a widespread strike affecting 185 karkhanas and more than 1600 workers for 21 days. This strike centered on the dearness allowance and perhaps on the reduction of the tobacco allowance.Footnote 108
But the nature of labour conflicts in the handloom industry of Sholapur was always of a different character than those taking place in the local mill industry. Labour leaders and karkhandars alike were generally participants in the Padmasali society of Sholapur and were usually prepared to negotiate settlements of conflict without extended battles. In some cases, the unions left the resolution of disputes to big men in the caste, persons who often had connections to the karkhandars.Footnote 109 The Lal Bavta Union itself served to address grievances of karkhandars against workers, such as non-payment of debt, as well as to deal with concerns that were purely those of wage labourers.Footnote 110 The Communists and their worker-followers, moreover, did not identify owners as class enemies to be opposed in all circumstances; rather karkhandars and their workers were at times perceived as jointly being members of a larger class oppressed by the colonial government and by big businessmen. Soon after the Communists fought for worker rights in 1940, they tried to organise the asamis and ordinary weavers alike to pressure the government to provide adequate supplies of yarn during the context of World War II. V.R. Madur recalled that they viewed the handloom industry as involving chains of exploitation rather than simple worker-capital conflicts. ‘Just as a small fish [the worker] is eaten by a big fish [the asami],’ they used to explain to the asamis, ‘the big fish is eaten by the biggest fish [the large capitalist].’ Even when they pressed asamis for wage increases, he explained, workers realized that their immediate bosses could not grant these increases without getting higher prices from merchants. The Communist Vithalrao Falmari (himself a master-weaver) even tried to organize karkhandars who gave work to the asami-employers.Footnote 111 The presence of Communist trade unions among handloom workers is thus not an indication of intense class conflict within the karkhanas.
Finally, collective activity did not sustain the commitments of workers all through Sholapur. After the upsurge of 1940–41, strikes were sporadic and usually confined to specific workshops.Footnote 112 Many workers recall that they never joined the union or participated in struggles against employers during this period. Ties of dependency, debt and patronage between labour and weaver-capitalists continued to promote accommodative responses that worked in a countervailing manner to horizontal solidarities and workers' resistances. As the Sholapur Handloom Weavers Enquiry Report suggested in 1948:
“The relations that exist between the weavers and their employers are very often of a personal character and naturally the weaver is unwilling to join hands with any force seeking to curb the power of the employers to whom he is always indebted, and sometimes related. This precludes the possibility of a general strike, and stands in the way of developing trade unionism on the principle of collective bargaining.Footnote 113
So even in Sholapur, where unions were strongest, it is difficult to speak of the formation of an exclusive proletarian identity. Throughout the Bombay Presidency, handloom weavers in karkhanas possessed many different affiliations, including ones that joined them with employers and with caste elites. Each workshop might employ workers with a range of different kinds of linkage to the karkhandar, from kinship and tight dependence to very loose, more ephemeral, perhaps antagonistic, ties. In short, a variety of lines of consciousness, some that may have been appeared ‘traditional’ and some that seemed new, all acquired their shape in the changing environment of the weaving towns.
Conclusion
This essay has shown that questions of labour were nearly as central a feature of handloom production as they were in large-scale industries. At the same time, it has demonstrated that understanding production relations though resort to the paradigm of proletarianization may be inappropriate. If the concept of proletarianization is to indicate merely the spread of a system of wage labour, it clearly has some applicability in a context where larger workshops did slowly become more prevalent than dependent households. But if it is meant to suggest a universal process by which impersonal forces of capitalism dissolve a traditional social order and compel workers to participate in a new one based upon radically different principles, the concept clearly has limitations when it is applied to the handloom industry of the Bombay Presidency. Both small households and larger workshops sustained the importance of the family economy, the value of rural connections, artisanal attitudes toward time and leisure, caste affiliations, and the role of debt, patronage and social deference. These features were reproduced not simply because of the hold of ‘tradition’ but also due to new features of a ‘small town capitalism’ that was emerging in the context of global industrialization.
Most notably, the ability of merchants and weaver-capitalists to reshape production relations was never total. The market for handloom cloth in western India did not provide an environment that could support the levels of investment necessary for merchants or karkhandars to develop sustained and firm control over labour. Capitalists needed to mobilize workers in order to satisfy sometimes intense pressures to fill specific consignments during periods of peak business. But their ability to remold the shape of the workplace was constrained by the fact that they could often not provide the work needed to keep all their employees busy throughout the year or to offer sufficient inducements to pull in significant numbers of new entrants into the handloom industry.
As a result, owners had to rely primarily on ‘caste weavers,’ that is, workers from communities already associated with weaving. Such weavers were positioned to check the most heavy-handed intrusions into their work practices. Their everyday actions and the threat of other actions often forced capitalists in the handloom industry to back off interference with these practices even as they made alternative efforts to gain control over the cloth that was being manufactured. In order to guarantee access to a supply of cheap labour when it was needed, they relied on rewards beyond wages, especially loans and patronage services. In effect, they created sets of workers who were dependent-clients as well as employees. The forms of their relationships to labour were many, reflecting a variety of different accommodations in different local contexts. Only a decentered narrative, reflecting the political interplay of a variety of different actors in a great number of places, can hope to capture how the specific characteristics of Indian capitalism took shape. What is common with other labour regimes that have been studied in South Asia and in Europe is the effort by capital to subordinate labour, the struggles of workers to resist full subjection to discipline from above and the asymmetries of power in of most of these struggles.
To some extent, this article confirms the arguments of Stephen Marglin, who in writing on Europe, sees eighteenth- and nineteenth-century changes in the organization of production as spawned by conflicts of the work place rather than as a necessary requirement of the adoption of machine technology.Footnote 114 But this study clearly cautions us against seeing capitalists as being always in a position to predetermine the outcomes of such conflicts, irrespective of the characteristics of their workforce or their markets. The handloom workshops showed no signs of developing into large factories or of adopting intense regimes of labour discipline. Later new, more brutal, systems of control would become associated with the powerloom units founded by handloom karkhandars in some of the weaving towns.Footnote 115 The shape of these units, however, would also be profoundly affected by employees' struggles. Responding to large collective actions by workers to enforce the labour laws during the 1950s and 1960s, powerloom owners partitioned their workshops, reducing them to sizes in which factory legislation no longer was applicable and in which unions found it difficult to gain a foothold. Today the powerloom industry, an outgrowth of small town handloom production, has become India's largest source of industrial employment. But it is built upon a framework of thousands of small units run by karkhandars who themselves often exist in dependent relations with larger merchant-capitalists.Footnote 116 This framework itself, I would argue, exists less because of some rationality inherent in small-scale production; rather it arose in the specific historical context of efforts by capitalists to contain and manage workers' resistance.
